Temperature is a well-known cause of degraded performance and/or failure in electronic devices. When vacuum tubes were in widespread use, the heat generate by such devices as color television receivers tended to raise internal temperatures to the point that various electronic elements such as resistors would “cook” and change value, and sockets for the vacuum tubes, if made from organic material such as phenolic, would char and in some cases disintegrate.
With the advent of semiconductors and other solid-state devices, the amount of heat produced by an electronic device tended to be less than in vacuum-tube equipment. However, while the amount of heat produced by a solid-state apparatus may be less per device than in a vacuum-tube apparatus, the number of solid-state devices, and their packing density, can be so great that temperature-related problems are exacerbated.
Various methods have been used to control the temperature of solid-state devices. In any situation in which undesired heat is produced by a device, the heat must be taken away by a low-thermal-resistance path. Since heat flows from a region at a given temperature only to a region at a lower temperature, the number of options for temperature abatement are limited. When discrete packaged semiconductors were in widespread use, a thermally conductive, large-surface-area “heat sink” could be attached to the case of the semiconductor to increase the effective surface area of the case in order to provide a low-thermal-resistance path from the high temperature semiconductor to lower-temperature ambient air. The art of “heat sinks” became elaborate, especially as applied to cased microcircuits such as microprocessors.
Another technique for helping to control the maximum temperature of semiconductors in a heat-producing circuit is to connect a flat portion of the case of the circuit by means of a thermally conductive path to a thermally conductive “cold plate.” Those skilled in the art know that a “cold plate” is not necessarily cold to the touch, and needs only to be cold relative to the item to be cooled. One such arrangement with a cold plate is described in U.S. Pat. No. 6,316,719, issued Nov. 13, 2001 in the name of Pluymers et al. Ideally, a single such circuit is to be temperature controlled, and the broad side of the circuit plate can be affixed to the cold plate, as by screws or adhesive.
In some cases, it is necessary to provide a relatively large number of heat-producing circuits within a defined region. Such a situation occurs in the case of active array antennas, where the operating frequency and antenna beam characteristics tend to determine the inter-circuit spacing of the associated active circuits. In some cases, the active circuits may be only low-noise amplifiers. In such cases, the heat density of the array of active elements may be relatively low. When, however, the active circuits include transmitter functions, as in the case of an active array antenna for a radar system, a large amount of heat may be generated within each active circuit, and the active circuits may be in close proximity. Perfect solutions to temperature control of the active circuits under such conditions have not been found.
One solution to the problems of temperature control in the high packing density, radar active array context is described in U.S. Pat. No. 6,469,671, issued Oct. 22, 2002 in the name of Pluymers et al. In this arrangement, a plurality of transmit-receive modules are arrayed on each of a plurality of generally planar heat-conducting baseplates of a line-replaceable unit. The heat-producing modules are mounted along an edge of the baseplates facing the antenna array. This arrangement is effective, but may be costly, and some embodiments may undesirably not have all the electronic devices on the line-replaceable unit, but on the underlying frame.
Another way to cool solid-state devices is by use of a coolant fluid which transfers heat by convection to a heat sink. Such a scheme is described in U.S. Pat. No. 5,013,997, issued May 7, 1991 in the name of Reese. This arrangement has the advantage of reducing stress on the heated element, but may not be useful in the case of line replaceable units. Another scheme for the use of a coolant fluid is described in U.S. Pat. No. 5,459,474, issued Oct. 17, 1995 in the name of Mattioli et al. In this arrangement, the electronic modules associated with the elements of an array antenna are mounted behind the array on slide-in carriers, so the electronic modules can be exposed. The electronic modules transfer heat by a thermally conductive path to a cold plate, and a flow of coolant fluid transfers heat away from a cold plate. This arrangement is effective in maintaining the cold plate temperature, but does not improve the transfer of heat from the heat-producing modules to the cold plate. The fluid connections may be difficult to connect and disconnect, and may leak under some conditions. U.S. patent application Ser. No. 10/640,445, filed Aug. 13, 2003 in the name of Pluymers which is still pending, describes a coolant fluid quick-connect arrangement which tends to reduce spillage under certain conditions.
An arrangement is described in U.S. Pat. No. 6,388,317, issued May 14, 2002 in the name of Reese, in which a flow of coolant is provided to support elements immediately underlying the solid-state chip which is the source of heat. This arrangement is very effective in heat transfer, but requires a fluid path for each chip, which in turn requires plumbing for each chip, and may be expensive. The fluid paths, if numerous, provide many locations at which leakage can occur.
Improved or alternative temperature control arrangements are desired.